1413 COOCH,
COOCH,
RC-S,
H
H
I
RC-S
I C-
D
C
The formation of 1 in the absence of acids is believed to result from dimerization of A. The acid catalysis of the formation of 1 indicates that A is protonated to give the cation D. D would be an effective trap for the nucleophilic carbene A giving protonated 1. The formation of 2 is probably the reaction of carbenes A and B. If carbene B reacted with carbon disulfide analogously to the known reactions of diarylcarbenes with carbon disulfide,' the result would be 3. All analytical and spectral properties of 3 are in agreement with this structural formulation. The reaction of carbon disulfide, hexafluoro-2-butyne, and acetic acid at 100" gave a 20% yield of 1 (R = CFs) and a 70% yield of 9. None of the expected 8 was obtained. The absence of 8 may be rationalized by its expected ease of solvolysis to give the cation, CF3C-S CF,C-S'
I@)-
CF,C-S
+
CF,CSCCF,
100'
i-PrN=C=N.i-Pr
,i-Pr CF,C-N'
II I
CF,C-CH
N' 'C(CH,)?
[cF,i-;-
i- Pr
12
-
CF,C-C
CF,C-S
I@)H CF,C-S
'OAc
zwitterion is formed, it must cyclize at rates faster than diffusion-controlled protonations, for acyclic compounds have never been isolated in carbon disulfide additions. Attempts were made to generate imidazolium carbenes by thermal 1,3 addition of carbodiimides to acetylenes. Hexafluoro-2-butyne and diisopropylcarbodiimide gave only 12 derived from a 1,2 addition. It is thought that the precursor of 12 is the amidine 13 which does not enjoy the normal resonance stabilization of amidines.
CF&-C
N '-
i- P r
+
8
13
H. D. Hartzler Contribution No. 1649, Central Research Department E. I. du Pont de Nemours and Company, Experimental Station Wilmington, Delaware I9898 Received December 6, I969
equilibration of cation and carbene, and the probably irreversible addition of carbene to acetylene. The formation of 9 suggests that carbenes B have nucleophilic as well as electrophilic character. A completely analogous result has been obtained with carbon disulfide, hexafluoro-2-butyne, and benzoic acid. Carbonyl sulfide gave a different reaction with acetylenes. When carbonyl sulfide was heated at 100" with ethyl propiolate or hexafluoro-2-butyne starting materials were recovered unchanged. With methanol present compounds 10 and 11 were formed. The acyclic zwitterion E is believed to be an intermediate. An analogous intermediate in the carbon disulfide
+
HC=C-COOEt
COS
=?
Alkylations and Acylations with 1,3-Dithiolium Carbenes Sir: The thermal reaction of carbon disulfide with acetylenes having at least one electron-attracting substituent has been shown' to generate 1,3-dithiolium carbenes which have been captured by a variety of electrophilic reagents. Another general reaction of 1,3-dithiolium carbenes has been found in the a-alkylation of aldehydes and ketones. The reaction of hexafluoro-2-butyne, carbon disulfide, and acetophenone at 100" for 6 days gave a 79% yield of 2-(benzoylmethyl)-4,5-bis(trifluoromethyl)-1,3dithiole (14). Under comparable conditions a 75 CF,C=CCF,
+
CS2
+
CGHjCOCH3
'C=i-COOEt
CFSC-S,
1
I+r'
,CH-CH,COCGH,
CF,C-S
L O
E
H\
/c=c
CH3OC-S
/H 'COOEt
1I 0
10
additions would rationalize the need for an electronattracting substituent on the acetylene. If such a (7) A. Schonberg, E. Frese, and K. Brosowski, Chem. Ber., 95, 3077 (1962).
14
yield of 2-(2-oxocyclohexyl)-4,5-bis(trifluoromethyl)1,3-dithiole (15) was obtained from cyclohexanone and a 53 yield of 2-(diacetylmethyl)-4,5-bis(trifluoromethyl)1,3-dithiole (16) was obtained from 2,4-pentanedione.
cF3f-sbH6 CF3C-S
CF3C-S,
1
CH-CH(COCH3)2
CF,C-S' 15
16
(1) H. D. Hartzler, J . Amer. Chem. Soc., 92, 1412 (1970).
Communications to the Editor
1414 Analogous products have been obtained from acetone and from 2-butanone in slightly lower yields. In these cases products 2 and 3 (R = CF# from the acetylene and carbon disulfide alone were also obtained. With 2-butanone, alkylation occurred exclusively at the methylene group. Aliphatic aldehydes such as propionaldehyde and isobutyraldehyde gave analogous products in 28 and 29 yield along with 2 and 3. Dimethyl acetylenedicarboxylate reacted similarly with carbon disulfide and aldehydes and ketones. Slightly higher temperatures (120-140”) were needed and yields have generally been lower. This reaction is not general for compounds with acidic hydrogen. Ethyl malonate and malononitrile were not alkylated. In these cases products 1, 2, and 3 were obtained. In none of these reactions have products derived from nucleophilic attack of the 1,3-dithiolium carbene at the carbonyl group been found. The absence of a general reaction for acidic hydrogen compounds indicates that a direct insertion of the carbene into a CH bond is an unlikely mechanism. Reaction of the carbene with the enol form of the aldehyde or ketone also appears unlikely because of the low concentration of enol under the essentially neutral reaction conditions. The most likely mechanism for the alkylation reaction is given in Scheme I. That the dithiolium carbene is generated under the reaction conditions is Scheme I CF,C=CCF,
+
CS,
--+
pp rg
COC6H5
I( CF3C-S
11
CF,C-S’
C-
18
OH I
HCEC-COOCH,
CF3CECCF3
+
+
CS,
CSp
+
+
c1
COHSOH
100”
dithiolium carbene, protonation to give the corresponding cation, and a Friedel-Crafts type alkylation of the phenol (or phenolate ion). In some cases the addition of acid (such as trifluoroacetic acid) is needed to suppress the addition of the phenol to the acetylene. The thermal carbon disulfide-acetylene reaction permits the investigation of a wide variety of reactions of a carbene which is probably unencumbered with solvent or other complexing materials. H. D. Hartzler Contribution No. 1650, Central Research Department E . I . du Pont de Nemours and Company, Experimental Station Wilmington, Delaware 19898 Received December 6, 1969
R
demonstrated by the fact that products derived from it and the acetylene are also obtained in many of the reactions. Although in many cases the carbene reacts as a nucleophile the carbon atom still has a vacant orbital and the potential for electrophilic reaction. Electrophilic attack at the carbonyl oxygen would lead to the enol ether. The initial dipolar ion need not be an antiaromatic species. The negative charge could overlap with the d orbitals of sulfur and not enter into the T system of the dithiolium ring. Under the neutral reaction conditions the more stable enol would be formed and this would lead to the orientation observed with 2-butanone and 2,4-pentanedione. The isomerization of the enol ether to the observed product could occur by either a homolytic or a heterolytic process. Hexafluoro-2-butyne reacted with carbon disulfide and benzaldehyde to give 17 and 18. Cinnamaldehyde reacted analogously as did other aromatic aldehydes. It is believed that 18 is formed from 17 by the type of reaction discussed above for aldehydes and ketones. (2) The numbering of compounds is the same as in ref 1.
I
The mechanism of formation of 17 is not known, but a likely possibility would be nucleophilic attack of the 1,3-dithiolium carbene at the carbonyl carbon followed by a hydride shift. Another type of alkylation reaction was observed with acetylenes, carbon disulfide, and phenols. This reaction evidently results from formation of the 1,3-
CH-0-C
CF,C-S
It
,CH-CO-C,HS 17
i
CF,C-S, CH-CR,COR
CF&- S, 1s, -CCF, FH-C, CF&-S S-CCF,
CF,C-S,
Stereochemistry of the Alkylation of CobaIt(1)
Sir: In the first reported alkylations of vitamin BlZs,lthe reactions were formulated as additions to a cobalt hydride. Later it was proposed that B,,, is a nucleophilic Co(1) species and the alkylations are sN2 processes.2 Recent evidence has shown that B12sand the related cobaloxime(1) react with alkylating agents via bimolecular p r o c e ~ s e s . ~The variation in rate upon changing the alkylating agent is similar to that of known s N 2 reactions. However, the mechanism cannot be considered established until the stereochemical fate of the reacting carbon is known. We wish to report that the reaction of cobaloxime(1) with typical alkylating agents (epoxides, bromides, and (I) (a) 0. Muller and G. Muller, Biochem. Z . , 336, 299 (1962); (b) 0. Muller and G. Muller, ibid., 337, 179 (1964), (c) A. W. Johnson, L. Mervyn, N. Shaw, and E. L. Smith, J. Chem. Soc., 4146 (1963). (2) G. N. Schrauzer, R. I. Windgassen, and J. Kohnle, Chem. Eer., 98; 3324 (1965). (3) G. N. Schrauzer and E. Deutsch, J. Am. Chem. Soc., 91, 3341 (1 9 69).
Journal of the American ChemicaI Society I 92:5 / March 11, 1970
